1. Start with pixel clusters from FlowSOM (Gaussian blur sigma=2).
  2. For each point, for each pixel, look at all neighboring pixels with a radius of 2, i.e. 5x5 kernel around the center pixel. Count the number of cluster j, then divide by number of pixels in cluster i (where i is the center pixel). Divide by the total number of cluster i.
  3. Average across all points.

Pixel clusters x marker

Breakdown of FlowSOM clusters

## 
##       1       2       3       4       5       6       7       8       9      10 
##   18139   37490    4120   22889   22499    4255   50174   10308   18780    3738 
##      11      12      13      14      15      16      17      18      19      20 
##   44483   33009   56265   39160   60519   25506   90123   34074    3193   27648 
##      21      22      23      24      25      26      27      28      29      30 
##   18393   60321   89526   95429   61008  177108  166256    6195   39653    2909 
##      31      32      33      34      35      36      37      38      39      40 
##    8090   45685   20763   98339  132159   22182   54429   17178   79123   22338 
##      41      42      43      44      45      46      47      48      49      50 
##   27248   53528   95038  140246  216850  136414  205682  393130   52612  237841 
##      51      52      53      54      55      56      57      58      59      60 
##   18700  100084   78248  155889  236098  360600  525779  163233  344001 2456314 
##      61      62      63      64      65      66      67      68      69      70 
##   15501   16856   35281  108048  145752  237147 1121175   17734  444896  194666 
##      71      72      73      74      75      76      77      78      79      80 
##   32303   17642   52553   36215   94348  220753  621535  137140   59561   62016 
##      81      82      83      84      85      86      87      88      89      90 
##   35059   11674   25856   74795  157503  268920   57016   47355   20280    5556 
##      91      92      93      94      95      96      97      98      99     100 
##    5674    5183   11985   23110   95948  130842    1800    2050   41978    1051

All points

Plot with range 0-1

## [1] "Ratios that are greater than 1"
## [1] "c6, clust_16: 1.46366107246651"
## [1] "c10, clust_20: 1.56856150039563"
## [1] "c30, clust_40: 1.1434552795048"
## [1] "c28, clust_49: 1.65932927786624"
## [1] "c42, clust_52: 1.09616864313587"
## [1] "c3, clust_61: 1.26488895626357"
## [1] "c91, clust_62: 1.0793190995471"
## [1] "c61, clust_63: 1.34610394995692"
## [1] "c97, clust_68: 1.06917519694082"
## [1] "c62, clust_71: 1.0500613632272"
## [1] "c74, clust_75: 1.86760531057867"
## [1] "c73, clust_84: 1.98850991112275"
## [1] "c100, clust_90: 1.07776118233406"
## [1] "c74, clust_95: 1.2232903313378"

Plot with range 0-0.5

Plot with range 0-95th percentile

Histogram of scores

All points, zscore

Each point separately